Continuous background monitoring of plant performance utilizing a single-chip microcomputer and PRTN sequences

The authors demonstrate a method suitable for a single-chip microcomputer or VLSI implementation that provides continuous real-time background monitoring of linear electromechanical systems. In this implementation method, pseudorandom noise is generated and digitized with a single-chip microcomputer and utilized to observe shifts in plant performance by monitoring the impulse response. A Butterworth filter was chosen to simulate the electromechanical system for ease and convenience of transfer function modification during testing. The feasibility of monitoring and detecting shifts in plant performance using pseudorandom noise in the background mode in real time while the plant continues to carry out routine control was demonstrated experimentally. Guidelines are provided for selecting the pseudorandom noise amplitude and the analog/digital quantization level. Pseudorandom trinary noise was demonstrated to be superior to pseudorandom binary noise.<
>